Overview

This high-speed computer vision library for ARM NEON based processors was developed by Uncanny Vision (Head Office: India). 

With changes to hardware completely unnecessary, systems can be made faster simply by replacing the currently used functions with the functions in this library.

Features

  •  • Super high-speed C language functions for computer vision
     • Optimized for ARM NEON (for ARM Cortex-A8/A9/A15/A7/A5 core processors based on ARM NEON)
     • Performance is 2 to 20 times faster than non-optimized code such as OpenCV
     • More than 70 function types (with more under development)
     • A free evaluation version is provided

Application fields

  •  • ADAS (Advanced Driver Assistance Systems)
     • Machine vision
     • Surveillance cameras
     • Computer games
     • AR (Augmented Reality)
     • Gesture recognition
     • Camera-based smartphone apps, wearable computing
     • Robots
 In addition, various other embedded computer vision systems

Optimized operations

  •  • Effective utilization of ARM NEON
     • Floating points rewritten to fixed points
     • Data minimization
     • Use of look-up tables
     • Elimination of data transfer
     • Data reuse
     • Use of integral image
     • Optimization at the algorithm level
     • Effective utilization of cache
     • Calculations based on fixed points

Benchmark examples

Algorithm
Throughput
(megapixel/second)
Speed increase compared to non-optimized code, such as OpenCV
Canny Edge detection
25
3 times
ORB(1500 key points)
3.7
5 times
Convolution filter 5x5
96
22 times
Dilate / Erosion1536.5 times
Integral Image filter962.4 times
Harris Corner Detection15.76.5 times
Fast9 Key point detection242 times
Face detection(LBP cascade)-3.5 times
Connected Components(Image Dependent)
1.7 times
Pedestrian detection using HOG1.7
(on ARM Cortex-A15)
9 times
*Measured on ARM Cortex-A9 1 GHz. (On ARM Cortex-A15 800 MHz for HOG only)

Examples of use in automotive systems

  •  •  Pedestrian Detection
     •  Object Detection using HOG and LBP cascade detectors
     •  Forward Collision Warning
     •  Stereo Disparity based Warning system
     •  Lane Departure Warning
     •  Car Reverse Warning
     •  Surround View Systems
     •  Blindspot detection

Algorithm examples

Pedestrian Detection Car Detection

Stereo Disparity

Lane Detection

Face Detection


Algorithm list

Algorithms - High Level 
Stereo Disparity
Pedestrian detection using HOG
Face Detection using LBP
Lane departure
Vehicle Detection
ORB
Lucas Kanade optical flow
Background Substraction
Tamper Detection
Stereo Disparity Post Processing
Dense Optical Flow
Face Post Estimation

Algorithms - Mid Level
Hough line detection
Lens Distortion Correction
Perspective transform
Connected Components
Integral Image
Non-Maximal Suppression
Kmeans clustering, based image segmentation
Harris Corner detection
Canny edge detection
Fast9 and Fast12 corner detection
Homography estimation

Algorithms - Low Level
Convolution kernels for different data types
Morphological operations - erosion, dilation
Image resizing
Histogram
Pyramid - Averaging, Gaussian
Array Multiplication
Sobel edge detection
Color conversion - RGB2YUV, RGB2HSV, RGB2LAB
Flip, Transpose
Table Lookup
Rotation - 90, 180, 270

Related Products

Uncanny Vision
Deep-learning library
Uncanny Vision
Deep-learning library evaluation kit
Uncanny Vision
Super high-speed computer vision library
for ARM NEON based processors
PUX image recognition software

Contact Us